Highly efficient carbon dioxide photoreduction guided by machine learning and first-principles calculation

Photocatalytic reduction of CO₂ to high-value carbon-based fuels holds tremendous potential in addressing the growing energy crisis. However, the high C=O bond energy of CO₂ molecules (750 kJ·mol^-1) makes it challenging to activate and reduce CO₂.

Therefore, the construction of photocatalysts with novel electron transfer pathways is meaningful. Compared with the traditional single electron transfer channel, the development of multi-electron channels based on layered materials has obvious advantages in the improvement of carrier transport. Nevertheless, the rational design of a desirable photocatalytic model for multi-electron channels with optimized parameters is quite challenging.

Recently, a study titled "Constructing dual electron transfer channels to accelerate CO₂ photoreduction guided by machine learning and first-principles calculation" was designed and led by Prof. Jizhou Jiang from Wuhan Institute of Technology, China.

This work combines first-principles calculating and machine learning to successfully predict and prepare a novel BiOBr-Bi-g-C₃N₄ sandwich structure with dual electron transport channels for photocatalytic CO₂ reduction. There are three main reasons for the favorable activity by the novel structure:

(1) the introduced g-C₃N₄ nanosheets demonstrate a similar energy level structure with BiOBr, which benefits for forming an electronic superposition state;

(2) the excited carriers can be efficient separation and transferred owing to the special double electron transfer channels;

(3) since the photo-generated carrier of BiOBr and g-C₃N₄ have different time decay behavior, a multi-timescale reaction mechanism for CO₂ reduction can be constructed to optimize the reaction pathway.

An enhanced photocatalytic performance of CO₂ reduction (43 μmol g^-1 h^-1) is received by the BiOBr-Bi-g-C₃N₄ quantum well structure. Five machine learning models were used to explore the linear law of the various influence factors on the efficiency of multi-electron channels. The mechanism of photocatalysis was investigated systematically.

The results were published in Chinese Journal of Catalysis.